In an LTE (Long Time Evolution, Long Time Evolution) system, by using a configurable TAT (Time Alignment Timer, time alignment timer), a UE (User Equipment, user equipment) controls a time for which the UE maintains uplink synchronization. Specifically, if the TAT is running, it is considered that the UE is in an uplink in-sync state; if the TAT expires, it is considered that the UE is in an uplink out-of-sync state.
In the prior art, if downlink data sent by an eNB (eNode B, base station) is received, the UE needs to send, to the eNB, a feedback corresponding to the downlink data, where the feedback is used to notify the eNB whether the UE correctly receives the downlink data. Because it is considered that the UE is in the uplink out-of-sync state after the TAT expires, if the UE receives the downlink data sent by the eNB to the UE, the UE cannot send the feedback corresponding to the downlink data. To enable the UE to send the feedback corresponding to the downlink data, generally, before sending the downlink data, the eNB notifies the UE to execute random access to obtain uplink synchronization. Afterward, the eNB sends the downlink data, and the UE sends the feedback corresponding to the downlink data after receiving the data.
Therefore, if the UE receives the downlink data after the TAT expires, the UE cannot send, to the eNB, the feedback corresponding to the downlink data; consequently, the eNB may retransmit the downlink data for multiple times, downlink data transmission resources between the eNB and the UE are wasted, and an additional signaling overhead and delay are also caused when the UE executes a random access procedure.